Peter Trosper

Data foundations for AI in customer service operations

As Head of Science and Analytics for Associate Technology and Innovation, Peter led the development of a standardized, self-service source of truth for customer service performance metrics, creating reliable data foundations for AI tooling. He directed teams building machine learning and rule-based coaching tools that reduced average handle time and improved associate performance. He kept associate data use compliant with GDPR and regional privacy rules across North America, Europe, and Asia.

Most recently, as a Principal BI Engineer and technical advisor, he led the measurement strategy for a next-generation customer service platform, defining its KPIs and operating standards. He built an interim analytics solution that queried data lake assets without ETL overhead, and replaced an error-prone query process with a table-driven model that was adopted across the broader customer service organization.

Building Amazon Shipping’s BI from scratch

Peter built and led the business intelligence and data engineering team that supported the incubation and launch of Amazon Shipping, Amazon’s small parcel delivery service for sellers. He hired and developed the team and built Amazon Shipping’s reporting infrastructure from the ground up, including Redshift clusters and Tableau-based self-service reporting. He introduced agile sprint prioritization to speed up delivery and integrated new product data streams into the enterprise data lake.

A founding member of Fulfillment by Amazon

Peter was part of the founding team that launched Fulfillment by Amazon, holding roles across finance, business analysis, and program management. He led initiatives to reduce seller contact rates and speed up issue resolution, and delivered global pricing recommendations through detailed cost analysis and forecasting. He joined Amazon as a senior financial analyst and controller after six years of banking analysis at Washington Mutual and Wachovia.